This is a Kalman filter, realized in Simulink for easier understanding purpose.
Although the characteristics of a Kalman filter are well documented and understandable after reading a few papers and books, I had quite a hard time understanding the information flow entirely. The main reason for that was the confusing notation to distinguish between predicted and measured values, as well as future and past values. Therefore I made a small Simulink Model, which should help explaining the information flow and understanding the equations of Kalman filtering.